Studying the Social Aspects of Leisure: Development of the Multiple-Method Field Investigation Model (MMFI)

Maureen Glancy Sandra L. Little Department of Recreation and Graduate School Leisure Studies Illinois State University San Jose State University

A Multiple-Method Field Investigation Model (MMFI) is proposed in order to provide a strategy for balancing social psychological inquiry of leisure. The Model extends former multimethod recommendations, being organized on two dimensions. These are (a) proximity or closeness to the interactive leisure context or incident (i.e., direcdy interacting, observing interaction, or using out- of-context records), and (b) time relative to the leisure context or incident (before, during, and after). A number of methods, including observation, interview, and primary and secondary records have been incorporated to suggest how time and proximity to the interactive context/incident can be operation- alized to examine the social factors of personal leisure more adequately. KEYWORDS: Multiple methods, leisure, social psychology, social interaction, meaning

What was stated over a decade ago still holds true: social psychological research on leisure is psychologically rich and socially impoverished (Neu- linger, 1980). Similar to what Senn (1989) and Solano (1989) said about the broad scope of the discipline, in leisure research, there is a great deal of literature on psychological concepts in social psychology. There is much to read about motivation and needs, satisfaction, attitude, subjective definitions of leisure, personality and individual differences, and crowding and social- carrying capacity (Iso-Ahola, 1988). Although there has been some research on social aspects of leisure, e.g., Fine (1987), Roadburg (1983), Scott (1991a, 1991b), Smith (1985), and Zurcher (1970) there is much less to read about the contexts and experiences of individuals who meet and interact. Fre- quently missed are the face-to-face process of constructing special meaning in leisure, the social systems people create which influence each other and their leisure experience, and the sense of mutuality expressed by small groups of related people who share a leisure experience. It makes sense that we should turn to social psychology to provide us with theories to study leisure for it is in the social setting of recreation that so many people find moments of happiness, hone skills for challenges worth doing, or develop relationships in shared experiences. Leisure has been re-

This article is based upon a paper presented at the Sixth Canadian Congress on Leisure Re- search, University of Waterloo, Waterloo, Ontario, Canada, May 9-12, 1990. Inquiries or requests for reprints may be addressed to either author as follows: Sandra L. Little, Graduate School, Illinois State University, Normal, IL 61790-5121, or Maureen Glancy, One Washington Square, San Jose State University, San Jose, CA 95192-0060.

305 306 GLANCY AND LITTLE ferred to as the prime social sphere in which people can make choices, meet, develop relationships (Cheek & Burch, 1976; Kelly, 1983) as well as freely pursue and experiment with knowledge and, in challenging uncertainty, find ways for enjoying personal growth (Csikszentmihalyi, 1990). It is in interacting with rules, roles, re-enactments, realizations, and rewards created by recreation opportunities that the psychological individual meets society on a personal basis and becomes a social person. Although the social group has been central to a number of studies in the past, it was the structures of interpersonal influence that were focal rather than the process of creating those social structures and personal and group products. Several examples can serve to illustrate this idea. Using survey research, influence of primary-group relationships on recreation choice- decisions was reported by Burch (1969), and West (1984) published a research note examining the power of interpersonal relationships to explain adoption of new outdoor recreation activities. Social group affiliation has been examined in a number of ways and found to relate to participation in outdoor water activities (Bryan, 1977; Buchanan, Christensen, & Burdge, 1981; Christensen, 1980; Field & O'Leary, 1973; Kelly, 1974, 1983). Pub culture was characterized as a form of social recreation for people of working class background by Smith (1985) who used participant observation for his study. The influence of secondary relationships allowed Stokowski (1990) to expand the concept of social relations to include interactions of individuals in wider social networks. Altogether, these and other studies do suggest the importance of social elements in people's lives and in recreation. Although illuminating basic dimensions of interaction, these studies do not provide us with an adequate understanding of the complex interrelationships and meanings which develop. More recently, the application of participant observation as a means for studying leisure experience has been noted through the use of the informal interview (Moeller, Mescher, More, & Schafer, 1980), qualitative structured interview (Howe, 1988), and the in-depth, unstructured interview (e.g., Hen- derson & Rannells, 1988). In these instances, there has been an association with interactive, participant-centered research roles which allow the researcher to become the willing subject-in-training in order to gather information which accurately reflects the participant's point of view. From these techniques, a sense of understanding has been emerging about the social influences which situational similarities and differences create in people's minds (e.g., Allison & Duncan, 1987; Bialeschki & Michener, 1994; Hender- son & Rannells, 1988; Robertson, 1994). These represent a significant de- parture from objective views of subjects and their leisure to subject-centered views of leisure in the subjects' way of thinking. There is a sense of authenticity because the language is that of the subject-as-participant. The picture presented is limited because authenticity and accuracy cannot be assured where these have been single-method studies. Emphasis on certain aspects of leisure and certain research methods has limited development and testing of social concepts to understand the dy- SOCIAL ASPECTS OF LEISURE 307 namic potential of re-creative opportunity (Hull, Stewart, & Yi, 1992) and leisure as a socially constructed phenomenon (Glancy, 1990). This imbalance has been noted by others. For example, Bultena and Field (1983) spoke of the need to study social systems associated with leisure, and Iso-Ahola (1980, p. 43) wrote that we lack "systematic theorizing regarding the relationship between social processes and leisure behavior." Thus, the role transformation process that goes on in recreation remained a mystery in the opinion of Gordon, Gaitz, and Scott (1978). Kelly (1981) noted that the sociology of leisure, too, had its limitations, being dominated by study of structural prop- erties of leisure experience to the detriment of learning about the dynamics of the interactive process. So far as we can see, there is little advancement on these concerns. To move toward resolution of this imbalanced social- psychological research perspective, we propose a Multiple-Method Field In- vestigation Model (MMFI) to study how leisure is shaped by people and the way it takes on meaning to those involved. Social psychology is a way of conceiving how individuals both effect, and are affected by, their experiences with each other over time. Within the social psychological framework are a number of theoretical approaches for understanding how the human being becomes part of social groups within the larger society. One often-cited, but little-examined, theory of how groups of people form and develop their own common bases for understanding leisure is that of social (or symbolic) interaction. This is the mental process which is used by people when they are involved in recreation experiences. In this sense, recreation is viewed as instrumental action which has the potential to become leisure. It is through the tangible acts of re-creation that the leisure state of mind is consciously acknowledged or realized (as perceived freedom and intrinsic reward). In essence, we say that people are able to understand and to communicate about themselves, others, and incidents around them through symbolic meanings they learn to attach to their recreational roles, places, things, and experiences. In very simple terms, a person can learn what horseback riding in the state park lands means to those who ride by talking to riders or by reading an article, letter, or poem about a rider's experience; comparing what is heard/read to what is already known; making mental adjustments to accommodate what has just been heard/read; and then mentally storing that idea for possible later use. Thus, we say the meaning of something is socially constructed because the person interacts with other people, with other forms of information, and/or with one's own memories in reaching an understanding. When we say that growth experiences, like leisure, are constitutive of one's self-concept, we mean that people lit- erally construct or develop themselves in the interactive process that goes on when they are involved mentally in their recreational experience (Csikszent- mihalyi, 1990; Kelly, 1983; Kelly & Godbey, 1992; Rossman, 1994). In critiquing the state of knowledge created with a bias toward psychological research in social psychology, Sherif (1977) offered an insight for leisure scientists to consider. Her statement indicated that social psycholo- gists may be driven by a desire "to be really acceptable to psychological 308 GLANCY AND LITTLE orthodoxy while neglecting contributions from other social disciplines on the scope and importance of the larger social environment" (p. 45). Thus, the fact that we know little about the interactive experience in which leisure meaning forms limits our capacity to understand leisure on a personal and empirical level. Sherif suggested that both the social and psychological views are necessary. It is possible to link the psychological and sociological perspectives. In this regard, several authors have provided important conceptual contributions about the constitutive potential of the human experience. For example, Pieper (1952) offered the idea that culture is the consequent of instances in which one's well-being, or divinity, is sensed and called that leisure. This is close to the psychological definition which refers to leisure as a personally constructed, subjective perception of well-being that may occur when one is engaged in an actual or imagined activity (Iso-Ahola, 1980). Leisure has also been portrayed as a non-rational social institution by Cheek and Burch (1976) who saw leisure evolving out of cultural phenomena just as Pieper described. Cheek and Burch explained that what people do to experience leisure often falls within a traditional set of alternatives that are interpreted as socially developed tastes. Integrating these various points of view about leisure allows us to say that leisure can be conceived of as socially constituted; changeable; and part of a collective, but hidden, cultural knowledge with the source being groups of individuals interacting and enjoying themselves. The psychological and sociological perspectives expressed here are not very far apart in that human experience is the basic building block of meaningful leisure in both views. What is central to our interest in improving empirical contributions to a social-psychologically informed knowledge of leisure is the need to learn about participants' roles and experience in coming to realize meaningful leisure. We need the capability to answer challenges that will affect future conceptualizations of leisure and the re-creative process. One example of concern is Lefkowitz's (1979) castigating prediction that, when people no longer create their own richly rewarding leisure experiences and perceptions, they will be at the mercy of professionals who will treat them like recreational rabbits lost in a "briar patch of alternatives for free time" (p. 395). Another example indicates the gap in our knowledge. First to write about programming recreation so people can use it to make their own leisure opportunity was Rossman (1989, p. 9) who clearly stated his perspective that "experiencing leisure, then, is something that individuals do—not something programmers do to individuals." However, the specific situational nature and conditions of this process remained unexplained. The gap between empirical reality and theory was characterized by Blu- mer (1954) as an area of vague or sensitizing concepts which could provide only a general sense of reference to the empirical instance. In other words, the researcher who is studying social interaction needs to be in the natural setting and develop concrete examples of the interactive experience through his/her own participation (Bruyn, 1966). Roadburg (1983) produced a study SOCIAL ASPECTS OF LEISURE 309 of soccer players and gardeners based on a combination of methods, first participant observation, then structured interviews. In the final analysis, he reported only the answers resulting from structured questions and these really only correspond vaguely to his subjects' actual experiences. So, although he spent a great deal of time with these players and gardeners, he was unable to authenticate how these people organized their leisure together, what the deeper levels of meaning were that they shared (e.g., Glancy, 1990), and how their interactions affected and demonstrated their particular understanding of work and leisure. To examine the idea that people perceive and define leisure situationally, Sherifs (1976) advice bears consideration: we need to do more than interpret their thoughts and their impressions as a matter of personal or aggregate historical record. Simply focusing on "self-centered acts and speeches [ignores] the effects of other people" (Solano, 1989, p. 37). There- fore, we must study the experience in which leisure meaning is formed, i.e., during the creative moment, the moment where the spirit and mind of the person(s) are most present and the person's understanding is shaped and reshaped as the situation and they themselves grow or change (Kelly, 1981). Thus, to know leisure, as someone else knows it, calls for the "scientific study of... actions by the individual (verbal and non-verbal) in concrete [specific] situations over time" (Sherif, 1976, p. 171). In other words, to put the social element into our social psychological research, it is necessary to engage in in-depth study and interpretation of specific, everyday leisure experience of interacting individuals as they engage themselves in their recreations.

Multiple-Method Field Studies in Leisure Settings Use of multiple methods to examine the interactive recreational experience is not a new idea. Small and large field studies have been conducted on the social group experience with a focus on the social organization of leisure. One investigation is a classic, three-year field study on conflict and cooperation among groups in summer camp (Sherif, 1976). Other studies include a softball social group (Glancy, 1986), personal goods auctions (Glancy, 1988a, 1988b), gun auctions (Olmstead, 1986), family camping for incarcerated mothers (Little & Stumbo, 1989), race track and gambling ca- sinos (Abt & Smith, 1982; Abt, Smith, & McGurrin, 1985), pool players, (Chick, Roberts, & Romney, 1991), and bridge players (Scott, 1991a, 1991b). Particularly salient was Little's (1985, 1988, 1989) study of conflict and cooperation among volunteers, utilizing a multiple-method approach in a four- year study of a major community arts festival. Another was a study on marina life which allowed Levy (1989) to propose that a sense of community can emerge in recreation settings where people interact and build ties over time. With each study, participant observation was a major methodology; however, interviews, sociographic and interaction analyses, sociometrics, secondary data analysis, observation, group appraisal, archive study, and surveys were among the overall variety of tools employed. It is out of experience with 310 CLANCY AND LITTLE

several field studies that the MMFI was conceived as an ordered approach to study the social experience of leisure.

Evolution of the MMFI Frustration with the limitations in social-psychological research on leisure and increasing professional dependence on universal models for recreation service provision and evaluation prompted analysis of the problem. Several conclusions dominated our thinking. For instance, it seemed that, among the active researchers in our field, there may have been an unwill- ingness to accept the full range of social psychological concepts with the result that group/social process was largely ignored. This would have been a reflection of Sherif s contention about the positivistic influence of psychological research. Also possible was the idea that an insufficient number of studies on interactive process and symbolic products are available to serve as exemplars. This would mean that critical mass may be a factor. Cost of con- ducting studies organized around face-to-face methodology would also be problematic. For whatever reason, the apparent consequent is that theory neither informs the field of practice to any notable extent nor does the field of practice inform theory. Thus, our approach was to take the essential concepts from social-psychological theory (how the individual becomes a social person through symbolic interaction over time) and to evaluate ways of using them in constructing a model to guide research in the field. Since personal development and growth is not static, one criterion that was essential to our model was its ability to focus on studying the dynamic possibility for creativity and change in interactive experiences. A second criterion was that the model be faithful to the idea that the person is an active player in creating and changing perceptions of him/herself, as well as the recreational situation. This meant the symbolic interaction tenet that the researcher's perspective be shaped by interacting directly with the subject(s) would also be incorporated. The final version of the model has evolved from (a) our study of social psychology and field research, (b) firsthand experiences in field research and programming and leadership in the field of practice, (c) teaching students to engage in field research and apply social psychological concepts in their professional practice, and (d) reviewer critiques of earlier versions of the model.

The Multiple-Method Field Investigation (MMFI) Model According to Sherif (1976), social psychological research belongs in the field, that is, the settings where people are naturally going about the activities which produce the incidents or phenomena of interest. Criticisms of social- psychological research focus on two points in particular: its reliance on self- report (primarily survey methodology) and the use of singular methods (Bur- gess, 1984; Sherif, 1976; Solano, 1989; Wylie, 1961). Due to the complexity of the natural, social situation, Sherif (1976, p. 90) advocated that "a com- SOCIAL ASPECTS OF LEISURE 311 bination of independent research methods is, in fact, needed in almost any social-psychological research," particularly in order to allow the investigator to view group life over time. To solve the complexity problem in recreation and tourism settings, Hartmann (1988) recommended the value of using a multiple-method approach. Burgess (1984) suggested that field studies which use multiple sources of data result in genuinely cumulative research. Fur- thermore, many have noted that research findings are strengthened by the support and confirmation made possible by a multiple-method approach where issues of validity and reliability are treated in the design (Brewer & Hunter, 1989; Bruyn, 1966; Burgess, 1984; Campbell & Fiske, 1959; Hart- mann, 1988; Miles & Huberman, 1994). Specifically, a multimethod approach is a way of "overcoming each method's weakness and limitations" and is a way of enlarging the scope of the study to include the structure, setting, and social processes (Brewer 8c Hunter, 1989, pp. 11-12).

Time and Proximity-to-Experience as Primary Dimensions In this article, we propose a two-dimension Multiple-Method Field In- vestigation Model to guide the selection of research methods for engaging in social-psychological research. We emphasize the idea that data collection methods must be selected which not only are likely to produce information appropriate to answering the research question, but also are suitable for use in the research context to provide data which are not altered by research objectives or activities. Our first premise is that the dimension of time is essential in plotting a multimethod social-psychological design. This responds to Sherif s time criterion and highlights the importance of assessing interactive experience and meaning across time. McGrath (1988) noted that social psychological research has neglected time as an important aspect of research activity. He stated that "time can enter social psychological research activities ... as a dimension of the behavior being studied, as a crucial feature of the study context, and as a basic parameter of the study design itself" (p. 7). More to the point, applying the time criterion to research is a way to uncover the dynamics of the recreation experience (Hull, Stewart, & Yi, 1992) and achieve certainty and confidence in results. The second premise guiding the model design is that proximity, or de- gree of investigator closeness, to the interactive experience responds to the need for authenticity and accuracy in data collection and interpretation. Filstead (1970) wrote that the human being is a symbol manipulator so, to know what is going on in someone's mind, the investigator must perceive and understand the symbol's significance to those involved in the specific situation. Therefore, we propose use of direct and indirect methods of data collection. This means that the researcher must seek natural ways to interact with the participant(s) being studied as well as to use other means of observation and/or artifacts of that situation to produce useful data. In addition, we adopted Burgess' (1984) suggestion that primary and secondary data 312 GLANCY AND LITTLE sources require definition so accuracy can be estimated when evaluating in- terpretations of field research findings. Primary data sources include both firsthand observation of the interacting person (s) and/or personal docu- ments, provided they are still in context. This means that several paragraphs from an extensive, four-week journal would not provide the researcher with adequate primary observation; whereas, the entire journal would be ideal as a primary data source. Secondary sources refer to data which are published or information which is taken out of context. Use of school grades, a book about the subject containing extensive firsthand documentation, or co-work- ers' observations would serve as secondary data; they can provide additional perspective on a situation but are not reflections of the specific, personal experience we seek to understand. The Proximity dimension in the MMFI ranges from the immediate situation, to a cluster of methods which are somewhat removed, and, at a third conceptual level, to the kind of methods which place the investigator in the least proximate position with regard to the person (s) or situation being studied. The three levels of proximity are: 1. Direct involvement, interacting with those involved in everyday leisure experience. 2. Direct, context-based observation of the leisure interaction or of expressive accounts by those involved. 3. Supporting, secondary data, including: observation of, or interaction with, participants outside the context/topic of the everyday leisure under study; supporting records which relate to the subjects or their leisure experience but which are taken or published out of context; and other useful literature. Confirming our approach in categorizing researcher proximity to data, Zelditch (1970) was found to have synthesized field information into three classes. These three types parallel ours. For example, observations or actions accompanied by meanings form what he called the Type I class. Observed qualities, actions, and frequencies of occurrences formed his Type II class. The Type III class had to do with records of rules, positions, members, and other historical data. In the first instance (methods of direct interaction in or about the leisure context), the researcher is a participant with a role in the interaction in which meaning is constructed and shared between people. Being privy to the perception-structuring process that people experience when engaged in a common activity challenges the researcher to find ways to mentally take the role of the leisure participant (Blumer, 1962; Glancy 1993). Roles range from co-participant to interviewer and depend on allowing the participant(s) to shape the researcher's way of thinking about the leisure experience being studied and shared (Glancy, 1993). In the second condition of data collection (methods of directly observing the interactive context or its mental replaying as the participant[s] un- SOCIAL ASPECTS OF LEISURE 313 derstands it), the researcher is limited to observation, imaginatively recon- structing the roles of the subjects being viewed based on what is seen, read, heard, or otherwise sensed but without benefit of interpretation or correc- tion by the subjects. These data can be used to support/reject developing theories; reveal behaviors, patterns, or variations not discernible when directly interacting with participants; or study verbatim details available in a recorded leisure sequence. Across time, observational records can stand as a series of snapshots which can answer questions about consistency and change in the situation, meanings, or the people involved. For the third and most removed perspective (methods which seek out- of-context, secondary support), the researcher collects objective data and other potentially informative records that do not explicitly reflect the interactive process or its products. Supporting methods, while not grounded ho- listically in the interactive incident, can provide psychological and social data that may assist in clarifying interpretation of some firsthand information gained from the interactive situation. From Bruyn (1966), we gathered that the researcher roles in a well- balanced study should bridge the full scope. This includes immersion as an interactive co-participant and interviewer, as well as performing observer roles and roles as an interpreter of quantified and naturalistic non-context records. Zelditch (1970, p. 217) asserted that "a field study is not a single method gathering a single kind of information." He argued that scope was determined by a variety of methods used to acquire adequate data efficiently. The idea is to gather appropriate data, from appropriate sources, and in appropriate ways to study the specific, everyday leisure situation. Since naturalistic investigations center findings on qualitative data that originate in the minds of the subjects under study, it is conventional that the research design emerge and develop as (a) the investigator becomes familiar with the participants and the situation in which leisure is experienced and (b) pre- liminary findings produce new questions to answer. Thus, it would be inap- propriate for a researcher to construct a fully representative design using the MMFI prior to beginning study. More appropriate is use of the model to challenge the potential for accurate and authentic understanding of the leisure experience during the investigation and to conceive of added methods which can improve the validity and certainty of findings. In this respect, we consider the ideal of balanced scope an important evaluative criterion. The focus of the MMFI is the everyday leisure experience as it occurs naturally. Kelly (1981) cited the leisure event as the central focus of study, not as a special occasion but as a typical or everyday kind of interactive episode. To refer to the interactive situation being studied directly or by in- depth interactions with leisure participants, the term Incident(s) has been used, adopting Zelditch's (1970, p. 218) interpretation that the simplest incident occurs "at the same time in the same place and a more complex configuration .... of the same type would be a sequence of incidents" which develops a history. Thus a recreation incident exists across time, whether it 314 GLANCY AND LITTLE be for a matter of minutes, a day, or as an intermittent or continuous sequence of incident.1 The actual duration of the incident(s) being researched is not an issue in this model; the important consideration is to continue the investigation until "theoretical saturation" (redundancy of data collected) occurs, assuring that all important aspects of the leisure experience have been studied sufficiently to produce all reasonable variations of the data possible (Glaser & Strauss, 1967, pp. 62-64). This means that briefly occurring or rare incidents are not likely to be good subjects for studying leisure meaning. Since they are more like encounters, the brief interactive occasion would create little opportunity for altering a person's thinking and way-of- being or for those involved to make lasting changes in the incident. Together, the concepts of researcher role proximity to the incident(s) (direct interaction, observation of interaction or interaction artifacts, and observation of supporting out-of-context records) and time (before, during, and after) form a two-dimension grid, Figure 1. The vertical dimension (Proximity) is used to identify researcher role-closeness to the individual (s) in action. The horizontal dimension (Time) suggests that data collection can occur continuously across time or at one, two, or more points in time in reference to the interactive Incident being studied. Although not typical of social research, we do encourage including pre- incident and post-incident times in the research if possible. In this way, the subjective experience of recreation and leisure can be viewed as: (a) an in- tervention to change something about the person (s) involved; (b) a means of creating or passing on culture, (c) or an indicator of something else. Collecting benchmark data on the individuals or group and developing ways to follow-up on personal change or meaning, well-after completing the experience, are typically employed in quasi-experimental research in leisure studies. Comparatively speaking, the gains to be made in adding the pre and post perspectives to field study research may be valuable in guiding and evaluating methods and in meeting standards of rigor in approaching questions of verification or validity of findings. From the Proximity and Time dimensions, a 3 x 3 grid was formed composed of nine cells as found in both Figures 1 and 2. Figure 2 adds illustrations of the kinds of methods which appear to respond to the conceptual character of each of the cells.

'Whereas leisure implies presence of a state-of-mind which enjoys a sense of freedom and realizes intrinsic reward, recreation Incident is used to refer to the socially interactive situation which mentally relates one or more individuals to objects and/or actions for the purpose of restoring energy or health or refreshing the mind or spirit. Interactions may take place mentally in reference to memories, thoughts, ideals, role models, or in reference to one or more persons involved in the same incident. Recreation is an interactive opportunity a person uses to achieve the leisure-state-of-mind. Thus, individuals who share a common recreational interest in different but comparable incidents may be studied as well as groups in which everyone shares the same incident. SOCIAL ASPECTS OF LEISURE 315

CLOSER 1 2 3 Pre-Event Event Post-Event Interaction Interaction Interaction Interaction

4 5 6 ^ Observation of Pre-Event Observed Post-Event ^ Interaction Observation Interaction Observation X 7 8 9 O Supporting Pre-Event Primary/ Post-Event P^ Records of Record Secondary Primary/ O-i Interaction Records Secondary I FURTHER Records t t t Before During After

— TIME — Figure 1. Multiple -Method Field Investigation Model: A nine cell grid based on dimensions of time and proximity to action.

Elaboration of Methods Used in the Model During Incident(s) Methodologies. In correspondence with our intent to emphasize the social aspects of social-psychological experience, the focus of the Multiple-Method Field Investigation Model is cell 2. In cell 2, methods are presented which can facilitate and document the shaping of the researcher's thinking to resemble that of the subject(s) for whom meaning has been conceived and has a life of its own. This means engaging in direct interaction with the subject(s) or informants until common understanding is achieved, such as: negotiating relations and information exchange with any situationally appropriate interactive means like conversation; dancing together; writing a play; designing an offensive strategy for the team; story- telling; buying and selling at an auction; exchanging letters; hiking together; or playing music together. Accepted interactive research methods are participant observation and the in-depth or probing, unstructured interview. While the participant observer is directly engaged in the recreation experience as a co-participant and trying to learn what it means to be a bona fide 316 CLANCY AND LITTLE

PRE-INCIDENT(S) DURING INCIDENT(S) POST-INCIDENT(S)

I CELL 1 METHODS CELL 2 METHODS CELL 3 METHODS N Focus Group Meeting Participant Observation Follow-Up In-Depth T Informational Meeting In-Depth Interviews Interviews E In-Depth Interviews Informant Feedback Incident Debriefing R Pre-Incident Training Interviews Meeting A Casual Conversations Informal Interviews Verification of Findings C Interaction in Other with Group T Social Situations Reunion Incident(s) I Personal Reflection V Narrative Diary E

O CELL 4 METHODS CELL 5 METHODS CELL 6 METHODS B Field Testing of Systematic Unobtrusive Continued S Subjects and/or Field Observation Interpretation of E Instruments with Field Notes; Video/Audio Tape or R Pre-Incident Sociographic Photograph Records V Unobtrusive Observation; Audio- of Incident(s) A Observation of Video Tape; Continued T Subjects or Setting Sequenced Photos. Interpretation of I Interpretation of Behavior Recording Participant Diaries or O Subjects' Application Interaction Analysis Other Creative N Narratives Interpretation of Products Related to A Participant Diaries or Group Incident (s) L Other Creative Continued Products Related to Interpretation of Group Incident (s) Written Records Observation of Participants in Other Social Settings

S CELL 7 METHODS CELL 8 METHODS CELL 9 METHODS U Structured Interviews: Group/Incident Retrospective Survey P w/Partidpants, Records Collection Post-Test Inventory P Experts, Others Self-Report Inventory Analysis O Social Survey Electronic Self Follow-up Structured R Paper & Pencil Pre-Test Monitoring; Time Interviews w/ T or Inventory Diary Participants, Experts I Existing Personal or Structured Interviews and Others N Group Records w/Participants, Archive Expansion and G Analysis Experts and Others Analysis Local Area Media Sociometric Analysis Information Repeated Test/ Comparable Inventory Biographical or Historical Narratives/Films Interpretation of Former Participant Accounts or Creative Products Figure 2. Examples of methods arrayed by time and proximity relations to the incident. SOCIAL ASPECTS OF LEISURE 317 member of the study group,2 the interviewer is a researcher who has the challenge of developing an open, unguarded relationship with subjects. As an interviewer, the researcher listens to information provided by the participant or participant-as-informant and interacts using questions and reflective comments to verify that his/her understanding of the situation corresponds to that of the participant. The researcher does not frame the field of response with pre-determined or leading questions and answers. Both research roles can be altered to include briefer and more focused informal or casual interviews, and both research roles are significantly strengthened by regular diary narratives reflecting on the personal experience of the interacting researcher. Altogether, it is this central column of methods (cells 2, 5, and 8) which permit more or less immediacy in studying the socially organizing processes in which people construct their own meaning while engaging recreatively. Cell 5 methods indicate roles in which the researcher simply observes the sequence of interaction in the incident as an outsider rather than being a partner in the social construction process. In these roles, observations can be systematic and consistent without influencing the natural sequence of interactions; however, there is no way to be sure that what the researcher thinks about his/her observations reflect the perspective the subjects are taking. Whereas unobtrusive observation implies a range of data-gathering techniques from archival research to "simple observations of the behavior of persons ... at play" using a variety of instrumentation means from physical equipment to physical traces (Denzin, 1978, p. 256), we delimited this term to non-interactive observation where the actual interactive sequence could be assessed due to the amount of data available. Thus, we suggest recorded or direct observation. This includes the use of videotapes, audiotapes, or systematically sequenced photography of the interpersonal action or its artifacts like a mural that was created over time, a collection of paintings or poems, or what may be salient aspects of the incident or setting. In essence, we found that our thinking was similar to Denzin's (1978, p. 265) five types of simple observations which are: (a) exterior physical signs of appearance of the participant or the place; (b) analysis of expressive movements ("ki- nesics" and "embodied posturing"); (c) physical locations (termed sociographic analysis by Glancy in 1990); (d) language; and (e) time-sampling, Figure 3. The advantage in using unobtrusive observation is that, insofar as emerging theories or concepts relate to language, group culture, patterns of action or relations, social systems, or group process, it may be possible to witness them unobtrusively. An example is the well-known use of observation by Whyte (1943) in his study of the corner boys. Systematic recording of socio-

2For discussion of recreation roles and group membership see Glancy, 1986, 1990, and 1993. Naturally occurring group roles are the participant observer's means for interacting. Roles will vary for most members and personal and/or situational limits to role performance may occur and must be acknowledged as they would be by any group member. 318 GLANCY AND LITTLE

Denzin: Exterior Physical Signs of Participant or Place

MMFI: Systematic Unobtrusive Field Observation Field Notes Behavior Recording Interpretation of Diaries or Creative Products

Denzin: Analysis of Expressive Movements

MMFI: Videotape Sequenced Photos Interaction Analysis

Denzin: Physical Locations

MMFI: Sociographic Analysis (Glancy, 1990)

Denzin: Language

MMFI: Audiotape, Videotape Field Notes

Denzin: Time Sampling

MMFI: Systematic Unobtrusive Observation Audiotape, Videotape Sequenced Photos Figure 3. Comparison of MMFI During-Incident Observational methods to Den- zin's five types of simple observation. graphic patterns, language, and other behaviors in this study helped shed light on emergent theories of the boys' social system and what gave them meaning in life. The behavioral records or transcripts created through use of systematic observational methods can be reviewed, possibly permitting additional theoretical insights later in the study period. We have avoided using the physical trace label for participant diaries and other creative artifacts that result from the leisure incident because of the behavioral perspective connoted; i.e., neither interaction nor quality of the relationship are observable in foot- prints, trash piled up, or other marks of wear and tear. However, physical trace observation would be consistent with the conceptual framework of cell 8 methods. In addition, methods, like those listed in cell 5, suggest being able to observe and analyze the interaction sequences or effects reflectively without the pressure of the face-to-face relationship. SOCIAL ASPECTS OF LEISURE 319

Methods listed in cell 8 suggest objective, out-of-context, and limited response data that may show a changing profile of the subject (s) and/or the incident (s) over time. These methods seek particular data to help answer particular questions that have emerged during the on-going analysis. Physical trace analysis and archive analysis are appropriate to this research perspective as are structured interviews and incident-related tests or self-report invento- ries that correspond to a formative program evaluation. Data collected by cell 8 methods also may support altering of the research approach being used during the incident as well as help confirm or deny certain findings emerging from cell 2 methods. Pre-Incident(s) Methodologies. Methods located elsewhere on the grid are useful to elaborate, co-relate, and verify findings. In general, the period prior to the recreation study-incident is still somewhat of a meaning-less opportunity but may provide the researcher with impressions about the disorderly mind, which being absorbed in the leisure experience, may alter over time (Csiksz- entmihalyi, 1990). Pre-incident methods (cells 1, 4, 7) offer views of the unformed group (e.g., pre-season meetings to introduce sport-league rules for team formation and play or for community theater try-outs; informational meetings for trips, hikes, courses, training, or potential members; structured or informal interviews with individuals appearing for registration, information, or screening). The valuable historical perspective begins with subjective, observational, and objective data that can be gathered early in the study. Direct interactions (cell 1) with persons who may be possible subjects yield opportunities to ground the researcher's thinking before the incident of significance occurs. In this way, the research question can be more clearly formed or confidently assessed. Investigator interaction at pre-incident organized meetings presumes researcher roles as participant (not leader due to the biasing potential) or as probing interviewer. To the extent that an extended period of time has inten'ened in the regular recreation routine of a group (summer break; off-season) and an informational meeting is held for new members or for continuing members, then this may be considered a type of pre-incident occasion. Where cell 1 focuses on firsthand interactions in which the subject(s) influences the social situation created in a focus group meeting, for example, cell 4 methods almost completely eliminate the opportunity for give and take between researcher and subject(s). Cell 4 methods encourage use of pre- incident^) observation and initial documentation of subject interaction. Op- portunities may include performance in field tests or other context-relevant situations, any narrative or expressive materials developed by the potential subjects relevant to the forthcoming incident, or unobtrusive observation of individuals if it seems likely to produce useful information. Cell 7 methods add further limitations by restricting communication to structured survey responses and utilizing secondary data and applicable social/historical information. Preference, interest, skill, or personal quality in- ventories; motive/purpose assessment; and demographic and biographic information may be useful to the study. In addition, beginning to read 320 GLANCY AND LITTLE

biographies, poetry, or letters of persons who may be comparable in some way to the intended subjects or well-founded historical accounts and news stories that present background data relevant to the subject(s) or the antic- ipated situation are also useful preparations. Reading can be continued for the duration of the study as questions and theories arise. Caution must be accorded the use of pre-incident(s) methods, however. In using pre-incident(s) methods, the investigator is actively interacting with persons who may or may not be the subjects during the actual incident situation and is interacting with these people in non-incident situations. Bias can also develop when reviewing documentation of past incidents or other secondary evidence, of existing literature, and of potentially relevant theories. This means that, without careful clearing of the mind when entering the actual interactive incident, the researcher can carry mental expectations formed during the pre-incident(s) preparations. Bruyn (1976) explained that in the process of interpreting data, higher-level concepts are grounded in the investigator's concrete, personal experience (pp. 32-33); therefore, pre-incident(s) experience and early closure on concepts during the incident are both potential sources of bias. The best guideline is that use of pre- incident(s) methods is intended for opening the researcher's mind rather than closing it. Post-Incident(s) Methodologies. In general, post-incident(s) methods do two things. First they attest to the veracity of the study, and second, they allow continuing reinterpretation, theoretical reconceptualization, and reevalua- tion of significance over time. Post-incident(s) interactive methods shown in cell 3 are essential to ensuring the validity and reliability of any qualitative study. This array of methods makes a field study credible because the investigator takes the precaution of going public with the findings, asking for verification, to as great an extent as possible, by those who were directly involved. Furthermore, it is also appropriate to follow-up the interactive incident to learn whether or not the incident had lasting effects in the minds of the subjects. It is recognized that the process of recording (e.g., shooting photos/videos, writing a journal, or painting a picture) during the incident interaction can (a) be part of the experience, (b) extend the experience through memory recall and personal reflection, and/or (c) be used for post- incident analysis. Thus, what appears as a uni-dimensional act is really multidimensional. That is, verification of the experience is not just a post-incident act (as in summative evaluation), but an act of the incident itself. Within cell 6 are methods which facilitate reexamination of data so empirical reliability and theoretical validity can be reasserted by researchers or new insights can be observed and findings revised. Generally, there is such a wealth of data created in field investigations that continued analysis is necessary since complete use of data in a single analysis may be humanly im- possible. Use of post-incident data collection strategies, cell 9, which survey the original subjects and their primary or secondary associates provide a different means to expand the time dimension so that findings can be verified or SOCIAL ASPECTS OF LEISURE 321 corrected. In addition, comparison and correlation to other studies and pop- ulations are ways of evaluating the significance of the naturalistic field investigation. Conceptually, cell 9 methods also suggest the next stage of action. This may be either providing evaluation information to recreation practitioners (Glaser & Strauss, 1967) or developing an experimental or quasi-experimental investigation which tests hypotheses based on the grounded theory which emerged from the field investigation (Babbie, 1992).

Advantage of the MMFI Model Our experience suggests that the MMFI Model can aid in maintaining the balanced study advocated by Bruyn (1966) because both subjective and objective perspectives are utilized even when insights emerge and alter the research process over time. Where Brewer and Hunter (1989) wrote about systematic synthesis of research style (i.e., field study, surveys, experiments, and nonreactive studies), our goal is a systematic use of methods appropriate to probing and elaborating the interactive social situation both intraperson- ally and interpersonally. Justification for this strategy is the same as that suggested by Brewer and Hunter (1989): the MMFI provides a framework to reduce threat of rival hypotheses which are inherent in single-method or non-systematic multiple-method designs. Whereas other multiple-method approaches focus on construct validity and reliability (Campbell & Fiske, 1959) or improving certainty of interpretation (Brewer & Hunter, 1989), the MMFI Model treats the additional need for subjective accuracy in the interactive experience by incorporating the situational dimensions of time and experience. In addition, veracity of findings is enhanced because the MMFI Model offers a framework which utilizes three of Patton's (1990) concepts of tri- angulation. One of Patton's purposes for using multiple methods for data collection is as a vehicle for comparative analysis of findings. Patton con- tended that, although differences are often likely to occur, reconciling those differences in the conclusions of the study rather than in subsequent studies is an advantage. Use of methods at all three levels of proximity to the incident experience and at various times are ways that the MMFI Model can help the researcher gather data useful for comparative analysis. Targeting multiple sources of information is a second approach advocated by Patton. This promises opportunities to corroborate and verify findings during the course of study. In applying the MMFI Model, multiple sources are tapped, e.g., (a) the group itself, (b) individual members, (c) observers, (d) archives, and (e) other quantitative sources of data. Patton's third approach is to adopt multiple perspectives in analyzing data. Use of a combination of quantitative and qualitative methods integrates deductive and inductive perspectives and the attendant use of theory testing and theory generation procedures. The MMFI, with a range of methods based upon time and proximity to the experience, encourages a design which blends quantitative and qualitative perspectives. 322 CLANCY AND LITTLE

Conclusions The proposed Multiple-Method Field Investigation Model extends former multimethod recommendations to provide a strategy for balancing social psychological inquiry of leisure. Just as scholars in the discipline of social psychology note failure to include the social aspect, leisure scientists too, have erred in not maintaining balance in social-psychological investigations of leisure. Leary (1989) identified the problem as a lack of self-monitoring. We are all responsible, as scholars, for the scope of research perspective. The challenge is one of studying the social aspects of leisure experience at the moment of conception and in the recreative situation in which it is conceived. The MMFI Model identifies methods which capture the essential concepts underlying social psychology to frame data-generating options to study the interactive situation. These include participant observation, in-depth interviews, informal interviews, informant feedback, and personal reflective diaries. Beyond this, depth of interpretation can be gained by adding direct- context observation methods and supporting out-of-context and secondary record analysis. Scope, or breadth of perspective, can be enhanced by collecting data both prior to the start and following the end of the regular interactive situation being studied. By applying the proximity-to-expeiience concept (i.e., interaction, non-interactive observation, and collecting out-of-context supporting data) with time (i.e., pre-, during, and post-incident data), social aspects which constitute leisure perception can be described, interpreted, and verified empirically. This is not to say, however, that we must now swing the pendulum from the study of psychological aspects of social psychology to the societal side. We should, however, move to a more balanced approach to our research, that which studies social interaction. As members of a community of scholars, we are all in positions to encourage, if not engage in, research which deals with the broad spectrum of inquiry.

References Abt, V., & Smith, J. F. (1982). Playing the game in mainstream America. In F. Manning (Ed.), Worlds of play: The proceedings of the 1981 Association for the Anthropological Study of Play. New York: Leisure Press. Abt, V, Smith, J. F., & McGurrin. M. C. (1985). Ritual, risk, and reward: A role analysis of race track and casino encounters, foumal of Gambling Behavior, 7(1), 64-75. Allison, M. X, & Duncan, M. C. (1987). Women, work, and leisure: The days of our lives. leisure Sciences, 9(3), 143-162. Babbie, E. (1992). The practice of social research. (Sixth edition). Belmont, CA: Wadsworth. Bialeschki, M. D., & Michener, S. (1994). Re-entering leisure: Transition within the role of motherhood. Journal of Leisure Research, 26(1), 57-74. Blumer, H. (1954). What is wrong with social theory. American Sociological Review, 19(1), 3-10. Blumer, H. (1962). Society as symbolic interaction. In A. R. Rose (Ed.), Human behavior and social processes: An interactionist approach (pp. 170-192). Boston: Houghton Mifflin. Brewer, J., & Hunter, A. (1989). Multimethod research: A synthesis of styles. Sage Library of Social Research, 175. Newbury Park, CA: Sage. Bryan, H. (1977). Leisure value systems and recreational specialization: The case of trout fish- ermen. Journal of Insure Research, 9(3). SOCIAL ASPECTS OF LEISURE 323

Bruyn, S. T. (1966). Tlie human perspective in sociology. Englewood Cliffs, NJ: Prentice-Hall. Buchanan, T., Christensen, J. E., & Burdge, R. J. (1981). Social groups and the meanings of outdoor recreation activities. Journal of Leisure Research, 13C&), 254-266. Burch, W. R., Jr. (1969). The social circles of leisure: Competing explanations. Journal of Leisure Research, 1(2), 125-148. Bultena, G. L., & Field, D. R. (1983, October). Social systems: A new agenda for leisure research. Paper presented at the NRPA-SPRE Leisure Research Symposium, Kansas City. Burgess, R. G. (1984). In the field: An introduction to field research. London: George Allen & Unwin. Campbell, D. X, & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait- multimethod matrix. Psychological Bulletin, 56, 81-105. Cheek, N., & Burch, W. (1976). The social organization of leisure in human society. New York: Harper & Row. Chick, G., Roberts, J. M., & Romney, A. K. (1991). Conflict and quitting in the Monday night pool league. leisure Sciences, 13(4), 295-308. Christensen, J. E. (1980). Rethinking "social groups as a basis for assessing participation in selected water activities." Journal of Leisim Research, 12(4), 246-256. Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York: Harper & Row. Denzin, N. K. (1978). The research act. New York: McGraw-Hill. Field, D. R., & O'Leary,J. T. (1973). Social groups as a basis for assessing participation in selected water activities. Journal of Leisure Research, 5(2), 16-25. Filstead, W. J. (1970). Introduction. In W. J. Filstead (Ed.), Qualitative methodology: Firsthand involvement with the social world (pp. 1-11). Chicago: Markham. Glancy, M. (1986). Participant observation in the recreation setting. Journal of Leisure Research, 18(2), 59-80. Glancy, M. (1988a, July). Under guises of proletarian respectability: Auction addicts pass. Paper presented at the Leisure Studies Association International Conference, Brighton, ENG. Glancy, M. (1988b). The play-world setting of the auction. Journal of Leisure Research, 20(2), 135- 153. Glancy, M. (1990). The social organization of role-taking at the auction. Leisure Sciences, 12 (4), 349-366. Glancy, M. (1993). Achieving intersubjectivity: The process of becoming the subject in leisure research. Leisure Studies, 12, 45-59. Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory. New York: Aldine de Gruyter. Gordon, C, Gaitz, C. M., & Scott, J. (1978). Leisure and lives: Personal expressivity across the life span. In R. H. Binstock & E. Shanas (Eds.), Handbook of aging and the social sciences (pp. 310-341). New York: Van Nostrand-Reinhold. Hartmann, R. (1988). Combining field research methods in tourism research. Journal of Tourism Research, 15(\), 88-105. Henderson, K. A., & Rannels, J. S. (1988). Farm women and the meaning of work and leisure: An oral history perspective. Leisure Sciences, 10(\), 41-50. Howe, C. Z. (1988). Using qualitative structured interviews in leisure research: illustrations from one case study. Journal of Leisure Research, 20(4), 305-323. Hull, R. B., Stewart, W. P., & Yi, Y. K. (1992). Experience patterns: Capturing the dynamic nature of a recreation experience. Journal of Leisure Research, 24(3), 240-252. Iso-Ahola, S. (1980). The social psychology of leisure and recreation. Dubuque, IA: Wm. C. Brown. Iso-Ahola, S. (1988). The social psychology of leisure: Past, present, and future research. In L. A. Barnett (Ed.), Research about leisure (pp. 75-93). Champaign, IL: Sagamore. Kelly, J. R. (1974). Socialization toward leisure: A developmental approach. Journal of Leisure Research, 6(3), 181-193. Kelly, J. R. (1981). Leisure interaction and the social dialectic. Social Forces, 60(2), 304-322. Kelly, J. R. (1983). Leisure identities and interactions. London: Allen & Unwin. 324 GLANCY AND LITTLE

Kelly, J. R., & Godbey, G. (1992). The sociology of leisure. Suite College, PA: Venture. Leary, M. (1989). The state of social psychology: Issues, themes, and controversies. Newbury Park, CA: Sage. Lefkowitz, B. (1979). Breaktime: Living without work in a nine to five world. New York: Hawthorn Books. Levy, L. (1989). Community in a recreational setting. Leisure Sciences, 11(4), 303-322. Little, S. (1985). Conflict resolution and decision-making among volunteer organizers of a community leisure event. Eugene, OR: Microform Publications. Little, S. (1988). A recreation programming model linking individual-level recreation behavior to actions of social groups. Journal of Park and Recreation Administration, 6(4), 90-102). Little, S. (1989). Conflict resolution processes for successful leisure service delivery. Journal of Park and Recreation Administration, 7(4), 31-41. Little, S., & Stumbo, N. (1989). Development of a multidimensional program evaluation model for a correctional setting. In L. McEvoy & D. Howard (Eds.), Abstracts from the 1989 Symposium on Leisure Research (p. 70). Alexandria VA: National Recreation and Park Association. McGrath, J. E. (1988). The social psychology of time: Nexo perspectives. Newbury Park, CA: Sage. Miles, M. B., & Huberman, A. M.(1994). Qualitative data analysis: An expanded sourcebook. Thou- sand Oaks, CA: Sage. Moeller, G. H., Mescher, M. A., More, T. A., & Schafer, E. L. (1980). The informational interview as a technique for recreation research. Journal of Leisure Research, 12(2), 174-182. Neulinger, J. (1980). Introduction. In S. Iso-Ahola (Ed.), Social psychological perspective on leisure and recreation (pp. 5-18). Springfield, IL: Charles Thomas. Olmstead,A. D. (1986). Wliat will you give me? Buying and selling at public auction. Paper presented at the Qualitative Research Conference, University of Waterloo, Waterloo, ONT. Patton, M. Q. (1990). Qualitative evaluation and research methods (2nd ed.). Newbury Park, CA: Sage. Pieper, J. (1952). Leisure - the basis of culture. New York: New American Library. Roadburg, A. (1983). Freedom and enjoyment: Disentangling perceived leisure. Journal of Leisure Research, 15(\), 15-38. Robertson, B. J. (1994). Leisure in the lives of male adolescents who engage in delinquent activity for excitement. Journal of Park and Recreation Administration, 12(4), 29-46. Rossman, R. (1989). Recreation programming: Designing leisure experiences. Champaign, IL: Saga- more. Rossman, R. (1994). Recreation programming: Designing leisure expeiiences, (2nd ed.). Champaign IL: Sagamore. Scott, D. (1991a). Post-mortems as extended leisure experiences. In C. Sylvester & L. Caldwell (Eds.), Abstracts from the 1991 Symposium on Leisure Research (p. 10). Alexandria, VA: National Recreation and Park Association. Scott, D. (1991b). The problematic nature of participation in contract bridge: A qualitative study of group-related constraints. Leisure Sciences, 13(4), 321-336. Senn, D. (1989). Myopic social psychology: An overemphasis on individualistic explanations of social behavior. In M. R. Leary (Ed.), The state of social psychology: Issues, themes, and controversies (pp. 45-52). Newberry Park, CA: Sage. Sherif, C. W. (1976). Orientation in social psychology. New York: Harper & Row. Sherif, C. W. (1977, October) Bias in psychology. Paper presented at the Women's Institute of Wisconsin, Inc. Conference: "The Prism of Sex: Toward an Equitable Pursuit of Knowledge," Madison, WI. Smith, M. A. (1985). A participant observer stud)' of a "rough" working-class pub. Leisure Sciences, 7(4), 298-306. SOCIAL ASPECTS OF LEISURE 325

Solano, C. (1989). Interactive perspective: Getting the social back into social psychology. In M. R. Lear)' (Ed.), The state of social psychology: Issues, themes, and controversies (pp. 35-44). New- bury Park, CA: Sage. Stokowski, P. A. (1990). Extending the social groups model: Social network analysis in recreation research. Leisure Sciences, 12(3), 251-264. West, P. C. (1984). Status differences and interpersonal influence in the adoption of outdoor recreation activities. Journal of leisure Research, 16(A), 350-354. Whyte, W. (1943). Street corner society. Chicago: The University of Chicago. Wylie, R. C. (1961). The self concept: A critical survey of pertinent research literature. Lincoln, NE: University of Nebraska Press. Zelditch, M., Jr. (1970). Some methodological problems of field studies. In W. J. Filstead (Ed.), Qualitative methodology: Firsthand involvement with the social world (pp. 217-231). Chicago: Mark- ham.